Added data cleaning code by Shreya and dataset information by Esther …

…and Research Question info by Anna to Checkpoint 1
COGS108 · Feb 24, 2024 · 531328c · 531328c
1 parent cb79baa
commit 531328c
Show file tree

Hide file tree

Showing 7 changed files with 1,419 additions and 59 deletions.
diff --git a/Cottagecore.png b/Cottagecore.png
diff --git a/DataCheckpoint_Group011_WI24.ipynb b/DataCheckpoint_Group011_WI24.ipynb
diff --git a/ProjectProposal_Group011_WI24.ipynb b/ProjectProposal_Group011_WI24.ipynb
@@ -31,10 +31,17 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "- What style of clothes correlate to higher sales and ratings throughout the seasons?\n",
-    "- We are considering using TikTok trends to predict clothing popularity (based on ratings) based on trending clothing in TikTok videos.\n",
+    "<b> What is the relationship between time and interest in various TikTok fashion trends? </b> <br>\n",
     "\n",
-    "\n"
+    "In Depth RQ: <br>\n",
+    "How can we predict the level of interest in various TikTok fashion trends across New York, California, and Texas using TikTok trend interest score data from January 2019 to December 2023 using a model that inputs the fashion trend name, monthly time data, and region to predict an interest rating between 0 - 100 and an associated label of 'low' (interest score < 25), 'rising' (interest score < 50), 'popular' (interest socre < 75), and 'trending (interest score > 75) for the specified input? <br>\n",
+    "\n",
+    "This question aims to develop a predictive model that evaluates the popularity of TikTok fashion trends in different regions and times, using a quantifiable interest rating system. The focus on a state-by-state analysis allows for a detailed understanding of regional preferences and trends overtime. <br>\n",
+    "\n",
+    "Interest Score/Interest Over Time Definition:  <br>\n",
+    "The \"interest score\" on Google Trends represents the relative popularity of a search query in a\n",
+    "specific region and time frame. It is indexed from 0 to 100, where 100 signifies the peak popularity\n",
+    "for the term. This score does not indicate the absolute search volume but rather shows the search term's popularity relative to the highest point on the chart for the given region and time. A higher score means more people are searching for that particular term at that time, while a lower score indicates lesser interest. The data is useful for identifying trends and understanding how interest in certain topics changes over time. <br>"
    ]
   },
   {

diff --git a/VintageThrift.png b/VintageThrift.png
diff --git a/Y2k.png b/Y2k.png
diff --git a/e-girl.png b/e-girl.png
diff --git a/fairycore.png b/fairycore.png